2.0 Analysis 2.1 Cause of Ferry Capsizing Because the ferry's lightship weight and draught have increased since the ferry was built, her effective freeboard is reduced. Although the ferry's initial transverse stability is relatively high, the range of positive stability is dependent upon the retention of adequate freeboard. Consequently, symmetrical athwartships and fore-and-aft distribution of vehicles is essential in preventing premature immersion of the main deck edge. In this instance, the trim associated with the positioning of the truck, in conjunction with the restraint on the hauling cable and the accumulation of flood water in the No.1 starboard side compartment, resulted in some two-degree starboard heel and a reduction in the after freeboard on the starboard side. In the prevailing 30cm-high waves, this condition permitted water to be shipped regularly on the after end of the main deck. As the moving truck gained backward momentum, the initial heel and trim would have tended to slew the truck slightly to starboard. With the vehicle only some 30cm off-centre, the starboard after corner of the main deck would be submerged some 20cm. With the added effect of the 30cm-high waves, almost 50percent of the aftermost main deck area would be suddenly inundated. (See AppendixA, Figure No.2.) The submersion of the after end of the main deck caused the centre of flotation of the remaining intact waterplane to suddenly move forward of midships and greatly increase the dynamic trimming effect of the moving vehicle. onsequently, as the ferry was settling rapidly by the stern, the dynamic trimming and slewing effects of the moving truck overcame the remaining longitudinal and transverse righting ability of the ferry, and the restraint of the hauling cable caused the ferry to capsize to starboard. 2.2 Vehicle Loading and Securing The truck was loaded on the centre line under the direction of the ferry operator, and was positioned such that the ferry was trimmed by the stern on departure. Fore-and-aft movement was restricted by fitting wooden chocks behind the rear wheels only--no other means of securing was employed or provided. The frictional contact between the truck and the wooden deck sheathing was reduced because of the truck's worn tires and of the wet and oily condition of the wood surface. In the absence of any arrangement to secure the vehicle to the deck, the lack of traction, in conjunction with the slight motion of the ferry, would have permitted the tires to slide on the sloping wet wood surface. Because the vehicle's ability to roll backward was mechanically limited and because the nature of the initial movement of the vehicle's wheels (rolling or sliding) was not noticed by either the ferry operator or the truck driver, it could not be determined whether the vehicle rode over, pushed back, or pushed aside the small wooden chocks fitted behind its rear wheels. In any event, the wooden chocks proved ineffective. 2.3 Ferry Operation and Operating Practices Owners' records included no documentation related to ferry stability, lightship weight or freeboard revisions due to machinery or structural changes, nor any formal operating or maximum deck loading instructions for the guidance of the operators. Consequently, the operating practices adopted--including deck load limitations and the decision of when to curtail ferry service because of adverse weather--were based on the accumulated on-the-job experience of the operators. 3.0 Conclusions 3.1 Findings The owners did not provide the operators with formal instructions or guidance regarding maximum deck loading or vessel trim and stability. A public notice showing the maximum deck load was not displayed at either of the ferry landing areas. There are no public highway weigh scales on Wolfe Island, nor is there any other suitable facility at the quarry. The determination of the laden weight of the vehicles carrying rocks between WolfeIsland and Simcoe Island is based on the driver's estimates and the ferry operator's observations. Shell plating damage causing flooding in a starboard side underdeck compartment was not reported to TC Marine, and remained unrepaired for several weeks before the occurrence. The cumulative effect of machinery and structural changes and additions since the ferry was built was an increased lightship weight and draught. The inherent effect of the hauling cable and the accumulation of flood water in the damaged compartment caused a slight list to starboard. The routinely adopted fore-and-aft location of trucks loaded on deck caused a significant trim by the stern, reduced the after freeboard, and made the ferry vulnerable to shipping water on the after end of the main deck. The frictional contact between the truck and the wooden deck sheathing was reduced because of the worn and damaged condition of the truck's tires, and of the wet and oily condition of the wood surface. The chocking arrangement adopted was ineffective in preventing the initial backward movement of the truck on the sloping wet surface of the wooden deck sheathing. The backward movement of the truck caused an increase in aft trim and immersion of the after end of the main deck, such that the ferry's longitudinal and transverse stability characteristics were suddenly markedly reduced. Due to the ferry's starboard list, the truck slewed slightly off the centre line as it moved backward, and this dynamic heeling effect overcame the ferry's remaining reserve stability. Trimming ballast located on the port side of the main deck to counter the weight and heeling effect of the starboard side machinery installation was unsecured. The bilges of the underdeck compartments were not routinely monitored. The buoyant life-saving apparatus was not fitted with a quick-release mechanism. The emergency boat and the buoyant apparatus were not fitted with retro-reflective tapes, nor were they painted with high visibility colour. The suddenness and rapidity of the capsizing precluded the ferry operator and the truck driver from donning lifejackets. 3.2 Causes The cable ferry SIMCOE ISLANDER, which was heeled slightly to starboard and trimmed by the stern on departure, capsized because longitudinal and transverse stability were lost when the vehicle loaded on deck shifted. The initial backward movement of the truck was induced by the ferry motion and poor frictional contact between the vehicle's worn tires and the sloping wet surface of the wooden deck sheathing, and was not arrested because of ineffective rear wheel chocking and the absence of any other securing arrangement. 4.0 Safety Action 4.1 Action Taken 4.1.1 Ferry Operation Following this occurrence, the Transportation Safety Board of Canada (TSB) and TransportCanada (TC) apprised the owners of safety deficiencies concerning the operation of the ferry. In addition, TC imposed specific requirements with respect to vessel operation prior to re-certification. Subsequently, the owners took several corrective measures, inter alia, restricting the maximum deck load, installing fittings to prevent the movement of the vehicles on deck, and establishing procedures to record daily sounding of underdeck compartments. Furthermore, the buoyant apparatus was fitted with a quick-release mechanism, and manhole covers were modified to make them watertight. A Ship Inspection Certificate (SIC)24 was re-issued by TC in September 1995 with a notation restricting the maximum deck load to 13tons pending submission of appropriate trim and stability data by the owners. In addition, TC issued a notice to other ferry owners reminding them of their responsibility to ensure that vessels are loaded and ballasted to maintain adequate stability, and to promptly report structural damage affecting the seaworthiness of their vessels to TC.